Functional characterization of the gonococcal polyphosphate pseudo-capsule

Neisseria gonorrhoeae is an exclusively human pathogen able to evade the host immune system through multiple mechanisms. Gonococci accumulate a large portion of phosphate moieties as polyphosphate (polyP) on the exterior of the cell. Although its polyanionic nature has suggested that it may form a protective shield on the cell surface, its role remains controversial. Taking advantage of a recombinant His-tagged polyP-binding protein, the presence of a polyP pseudo-capsule in gonococcus was demonstrated. Interestingly, the polyP pseudo-capsule was found to be present in specific strains only. To investigate its putative role in host immune evasion mechanisms, such as resistance to serum bactericidal activity, antimicrobial peptides and phagocytosis, the enzymes involved in polyP metabolism were genetically deleted, generating mutants with altered polyP external content. The mutants with lower polyP content on their surface compared to the wild-type strains, became sensitive to complement-mediated killing in presence of normal human serum. Conversely, naturally serum sensitive strains that did not display a significant polyP pseudo-capsule became resistant to complement in the presence of exogenous polyP. The presence of polyP pseudo-capsule was also critical in the protection from antibacterial activity of cationic antimicrobial peptide, such as cathelicidin LL-37. Results showed that the minimum bactericidal concentration was lower in strains lacking polyP than in those harboring the pseudo-capsule. Data referring to phagocytic killing resistance, assessed by using neutrophil-like cells, showed a significant decrease in viability of mutants lacking polyP on their cell surface in comparison to the wild-type strain. The addition of exogenous polyP overturned the killing phenotype of sensitive strains suggesting that gonococcus could exploit environmental polyP to survive to complement-mediated, cathelicidin and intracellular killing. Taken together, data presented here indicate an essential role of the polyP pseudo-capsule in the gonococcal pathogenesis, opening new perspective on gonococcal biology and more effective treatments.


Introduction
The rising emergency of antibiotic resistance and the lack of alternative treatments make of gonorrhea one of the major global health concerns with an incidence of over 87 million cases per year [1]. The complexity of the gonococcal infection occurring differently in the male and female genital tracts [2], along with a plethora of immune evasion mechanisms, impairing host defenses and preventing bacterial clearance, hinder the development of effective vaccine and new therapeutic approaches. It therefore becomes increasingly important to deeply investigate gonococcal pathogenesis since, despite significant efforts, this complex process is still characterized by unclarified mechanisms [3][4][5][6]. Gonococcus is described as unencapsulated bacteria, but its close relation with Neisseria meningitidis has often suggested the existence of a polysaccharide capsule. However, the absence of the region A in the capsule cluster of N. gonorrhoeae, responsible for the polysaccharide synthesis, precludes such possibility [7]. To this purpose, several groups attempted to demonstrate the evidence of capsular material through electron microscopy, producing inconsistent results [7][8][9][10][11][12]. Noegel and Gotschlich in 1983, isolated high molecular weight polyphosphate (polyP) located outside the cytoplasmic membrane of N. gonorrhoeae and proposed a possible role as capsule-like material [13]. However due the lack of proper techniques and controls, the existence of a capsule-like material in gonococcus remained controversial [13] and no further research to demonstrate the role of external polyP as pseudo-capsule in gonococcus was described in literature so far. PolyP is a polymer formed by hundreds of thousands of orthophosphates residues, linked by high-energy phosphoanhydride bonds and negatively charged at neutral pH [14]. PolyP has been found to be ubiquitous in all living cells, from bacteria to higher eukaryotes [15][16][17][18], occurring in subcellular metachromatic organelles known as volutin and playing an important role in the energy storage [19]. PolyP metabolism is regulated by different enzymes. In particular, its biosynthesis is primarily mediated by the polyphosphate kinase (PPK1), catalyzing the transfer of the γ-phosphate group of ATP into a growing polyP chain [20]. An additional polyphosphate kinase (PPK2), has been reported in some bacterial species, including gonococcus [21]. PPK2 differs from PPK1 because it preferentially uses GTP and Mn 2+ , rather than ATP and Mg 2+ , to catalyze the polymerization of polyP [22]. The catabolic reactions involving polyP are performed by both the endo-polyphosphatase and the exo-polyphosphatase (PPX) [23]. Despite polyP has been described to act as an energy storage polymer [24], numerous findings pointed out the importance of polyP also in biofilm formation, motility, sporulation, surface attachment, stationary phase and stress survival, dormancy and other features linked to the virulence of pathogenic bacteria like Pseudomonas aeruginosa, Campylobacter jejuni and Helicobacter pylori. [16,17,[25][26][27][28][29]. Moreover, in Neisseria meningitidis polyP has been proposed to be involved in the resistance to the bactericidal activity of normal human serum (NHS) [27,30].
In this study we unravel the existence of a polyP pseudo-capsule in N. gonorrhoeae and propose its role in gonococcal pathogenesis, specifically in the resistance to killing mediated by human complement, antimicrobial peptides and phagocytosis.

Transmission electron and confocal microscopy reveal the presence of a polyP pseudo-capsule in gonococcus
To allow transmission electron microscopy (TEM) analysis on the PorB.1B Neisseria gonorrhoeae FA1090 reference strain, the combination of the cationic dye, Alcian Blue, and the diamine L-lysine acetate, a specific fixative for labile material, was used. As shown in Fig 1A, this procedure of bacterial staining revealed the presence of a capsule-like material surrounding bacterial cells. The use of a cationic dye for the staining suggested the presence of negatively charged material, characterized by different levels of thickness and appearing better assembled in specific parts of the membrane. By omitting the fixative, this structure was not preserved, and it could not be visualized on the gonococcal surface ( Fig 1B).
To investigate the possibility that this capsule-like material could be related to external polyphosphate, the C-terminal polyP binding domain of the E. coli exopolyphosphatase (EcPPXc), was produced and purified as recombinant His-tagged protein to be used as probe to visualize polyP with confocal microscopy [31]. A clear and strong signal was obtained from the binding of EcPPXc on the surface of N. gonorrhoeae FA1090, thereby corroborating the hypothesis that the labile negatively charged capsule-like material observed with TEM was composed by polyP (Fig 2A). Localized spots were detected on the surface of a subset of bacterial cells (Fig 2B and  2C) and in some instances a marked high density signal on the septum ring was also seen ( Fig  2D). Furthermore, the analysis highlighted the presence of a polyP matrix embedding bacterial aggregates (Fig 2E and 2F). To further verify the specificity of this staining, free soluble polyP was added as a competitor for the binding of EcPPXc. Interestingly, no signal was observed upon competition ( Fig 2G) undoubtedly proving that the signal was specifically attributable to polyP.

The pseudo-capsule is differently allocated on the surface of gonococcal strains
Results reported in Figs 1 and 2 were also confirmed by flow cytometry analyses performed on N. gonorrhoeae FA1090 strain. Indeed, results revealed the presence of two distinct subpopulations, one showing a positive fluorescence shift due to the EcPPXc binding to polyP and a second, unreactive to EcPPXc, with no change in fluorescence (Fig 3A). To better evaluate the bacterial response occurring upon alterations of polyP homeostasis, single knockout mutants of each polyphosphate kinase (ppk) and exopolyphosphatase (ppx) genes as well as double knockout of ppk1 and ppk2 genes were generated in FA1090 strain. Flow cytometry analysis indicated that mutants lacking one or both polyP kinases exhibited a strong decrease in the population with surface-associated polyP (Fig 3B, 3C and 3D), whilst the exopolyphosphatase mutant resulted in a single population positive to polyP staining ( Fig 3E).
To investigate if external polyP was ubiquitously present among different N. gonorrhoeae strains, the study was extended to low-passaged clinical isolates. To this purpose, the serum resistant N. gonorrhoeae BG27, isolated in 2013 and expressing the same PorB.1B variant of FA1090 strain, as well as the knockout ppk1 mutant, were analyzed. Results of flow cytometry ( Fig 4A) and confocal microscopy (Fig 4B), revealed the presence of a single positive population characterizing the parental BG27 strain and an overall decrease in surface-associated polyP in the deletion mutant BG27 Δppk1.

Surface associated polyP content affects gonococcal complement-mediated serum resistance
To explore whether the polyP pseudo-capsule may be involved in gonococcal pathogenesis and specifically in the evasion from the complement-mediated bactericidal activity of human serum, the parental strain FA1090 and its ppx and ppk isogenic mutants were incubated with 10% and 20% NHS ( Fig 5A). Interestingly, while in the presence of 10% NHS the wild-type strain and ppx mutant strains survival was not affected, the three ppk mutants were significantly more sensitive to NHS. When the strains were incubated with 20% NHS, the wildtype strain was significantly affected and a further decrease of the survival rate of ppk mutants was observed. Conversely, the ppx mutant remained still unaltered, even when using this higher concentration of NHS, remarkably representing the most resistant strain. This suggests that the observed sensitivity of FA1090 wild-type to high NHS concentration may be due to the peculiar nature of this strain composed of two sub-populations without and with polyP exposed on their surface (Fig 3A). Based on these results, we could hypothesize a possibly different complement deposition mechanism on the surface of the different strains tested that may be inversely correlated with the polyP content on their surface. In order to support this hypothesis, we analyzed by flow cytometry the deposition of the C9

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule terminal complement component on the surface of gonococci after incubation with NHS. As expected, deposition of C9 on FA1090 Δppk mutants was significantly higher when compared with wild-type FA1090 whereas a negligible level of deposition was observed on the surface of FA1090 Δppx (Fig 5A).
Similar results were obtained when analyzing the survival of BG27 strain to NHS exposure ( Fig 5B). As it occurred for FA1090 strain, the survival of BG27 Δppk1 was hampered when using both 10% and 20% NHS. Interestingly, the wild-type BG27 strain was resistant to complement-mediated killing also at the highest serum concentration tested, suggesting that the homogenous presence of polyP in this strain may correlate with a wider protection from complement. These results correlated with the activation of the complement cascade as already shown for FA1090. Indeed, no deposition of C9 was observed for the parental strain, while a significant increase in complement protein levels was detected on the surface of BG27 Δppk1 strain ( Fig 5B). To visualize the polyP pseudo-capsule, flow cytometry analyses (A) and confocal microscopy (B) were performed on the clinical isolate BG27 and BG27 Δppk1. A. The grey filled profile represents the negative control stained with the fluorescent secondary antibody only, whereas the green areas represent bacteria with polyP pseudo-capsule, indirectly stained taking advantage of the His-tag of the recombinant EcPPXc. B. Immunofluorescence images show bacterial nucleoid in blue stained with DAPI and EcPPXc binding to polyP in green. Scale bar = 5 μm. https://doi.org/10.1371/journal.ppat.1011400.g004

The serum sensitive F62 strain naturally lacks a polyP pseudo-capsule
To corroborate the evidence of the role of polyP as one of the mechanisms of gonococcal serum resistance, serum sensitive strains were investigated, specifically the laboratory strain N. gonorrhoeae F62 [32]. When incubated in the presence of 10% NHS, F62 strain resulted to be sensitive to complement killing ( Fig 6A) in agreement with result that no polyP was detected on its surface ( Fig 6B). Interestingly, when exogenous free polyP was added to F62 culture, it could be detected by flow cytometry, indicating its binding capability to the bacterial surface ( Fig 6C). Given this evidence, we incubated F62 strain with exogenous polyP and tested its capability to survive when exposed to NHS. The results, in line with what observed for strains with naturally surface-exposed polyP, highlighted again the role of polyP as a shield to protect gonococcal surface from complement deposition. Indeed the presence of polyP correlated with F62 survival in NHS (Fig 6A). This ability may be considered polyP-specific and not purely dependent on local electric charge, since the addition of a different negatively-charged molecule like DNA, although coating the bacterial surface (S1 Fig), did not abrogate the NHS killing ( Fig 6A). Since gonococci in vivo may have a sialylated lipooligosaccharide (LOS) and that the addition of cytidine 5'-monophospho-N-acetylneuraminic acid (CMP-NANA) is able to induce LOS sialylation in vitro and increase serum resistance [33,34], we evaluated whether sialylation could interfere with polyP surface association. To this end, F62 strain was grown in the presence of increasing concentration of CMP-NANA (from 0.5 μg/mL to 500 μg/mL) and tested for serum resistance in the presence of NHS from two different donors. Results indicated that LOS sialylation increased survival to different extents, depending on the (B) were incubated with 10% and 20% NHS. In the case of FA1090 the wild-type strain, the single Δppk1, Δppk2 and the double Δppk1Δppk2 mutants were tested, while in the case of BG27, the wild-type and the Δppk1 mutant were used. Results of serum resistance assays are specified as percentage of survival calculated as the ratio of colony forming unit (cfu) after incubation with NHS over cfu incubated with heat inactivated NHS. Data reported are representative of 3 biological replicates performed with 3 different lots of NHS. Standard deviations are represented by vertical bars (*p < 0.05, **p < 0.005, ***p < 0.001). The ability of the strains to trigger complement activation was measured by flow cytometry after incubation of bacteria with 10% NHS. The difference of the median fluorescence intensity (ΔMFI) is calculated by subtracting the MFI in presence of heat-inactivated NHS from the MFI relative to NHS and shows the bacterial population reacting with human anti-C9 antibody. Complement fixation results are representative of one biological replicates performed with one of the NHS lots tested. https://doi.org/10.1371/journal.ppat.1011400.g005

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule complement source used (Fig 7). Results depicted in Fig 7A indicate that either CMP-NANA or polyP availability can confer full serum resistance to the tested strain. Consequently, no significant improvement of survival rate was measured testing polyP in combination with CMP-NANA. Noticeably, using a different complement source, CMP-NANA only partially improved F62 survival also at the highest concentration tested, while the addition of exogenous polyP totally restored serum resistance in both sialylated and not-sialylated bacteria ( Fig 7B). The transition from a serum sensitive to a serum resistant phenotype in the presence of exogenous polyP may be explained by the ability of polyP to hinder complement cascade activation as shown in Fig 6D, where it is demonstrated that the addition of exogenous polyP abrogated C3 deposition on F62 surface after exposure to NHS.

PolyP pseudo-capsule confers resistance to cationic antimicrobial peptide
Given the evidence suggesting a protective role of the gonococcal polyP pseudo-capsule against complement deposition, we questioned whether it could also act as a defense mechanism against antimicrobial peptides. To address this question, we evaluated the bactericidal activity of cathelicidin LL-37 against both the clinical isolate BG27 and its isogenic mutant Δppk1, previously characterized to have a reduced level of polyP surface exposure (Fig 4). Data revealed (Fig 8A) that the minimum bactericidal concentration (MBC) of LL-37 was four times higher for the parental strain compared to the isogenic Δppk1 mutant (Fig 8A, 2.5 μM vs 0.625 μM, respectively). The addition of exogenous polyP remarkably reduced the LL-37 sensitivity of the

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule mutant strain, with an MBC above the studied concentration range. On the contrary, when BG27 Δppk1 strain was pre-incubated with exogenous DNA, as a negatively charged molecule to mimic the potential charge effect associated with polyP, no effect on bacterial survival was measured, confirming that the protection observed was specifically linked to the presence of polyP rather than to its negative charges.
This analysis was extended to the polyP-negative PorB.1B F62 and PorB.1A SK92-679 strains (Fig 8B and 8C). Both strains displayed the same profile after incubation with increasing concentration of LL-37, showing an MBC of 1.25 μM. Hence, data depicted in Fig 8, likely suggested a direct implication of the polyP in the prevention of LL-37 mediated bactericidal activity as confirmed by experiments where exogenous polyP was included. Indeed, killing of SK92-679 strain even at the highest concentration tested, was totally abolished after 1 h of incubation, instead for F62 strain a slight decrease of survival was already observable with 2.5 μM of LL-37. Resistance of F62 to LL-37 was also tested after growth in the presence of CMP-NANA, the LOS sialylation agent conferring resistance to cationic antimicrobial peptide [35]. The major contribution of polyP than that afforded by sialylation in the protection from the cathelicidin LL-37 was clearly found and was even more evident under stressed conditions such as an increased incubation time of 2 h (S2 Fig).
Gonococcus survival in the presence of defensins was also investigated. In particular, we tested gonococcal susceptibility to α-defensins 1, also known as Human Neutrophil Peptide (HNP-1), mainly present in neutrophil granules. The HNP-1 bactericidal activity was evaluated on both the clinical isolate BG27 and its polyP deficient mutant, resulting in no impairment of the gonococcal survival at each concentration tested (S3A

The polyP pseudo-capsule protects N. gonorrhoeae from HL60-mediated killing
The potential involvement of the polyP pseudo-capsule in the gonococcal interaction with neutrophil-like cells was also evaluated. For this purpose, differentiated HL60 cells (dHL60) were infected with the N. gonorrhoeae clinical isolate BG27 and the isogenic mutant BG27 Δppk1 for 1 h. After previous removal of un-ingested bacteria, dHL60 were lysed and released bacteria were plated and then counted. Results suggested that the viability of the parental BG27 strain was significantly higher than that of BG27 Δppk1 mutant. Interestingly, the viability of the deletion mutant phenotype was reverted after free soluble polyP addition to the reaction. This suggests the involvement of the polyP pseudo-capsule in the increased survival of phagocytized bacteria (Fig 9A). To better understand if this phenotype was linked to a lower degree of bacteria internalization or to a decreased survival capability within dHL60, we developed a methodology to measure the amount of phagocytized bacteria by using an Opera Phenix system. Specifically, after infection for 1 h of dHL60 with 488-labelled bacteria, an antigonococcal primary antibody coupled with an AlexaFluor546 secondary antibody was used. This allowed to distinguish between internalized (488-positive only) or cell-associated (488/ 546-double positive) bacteria. As shown in Fig 9B, despite the number of dHL60-associated bacteria was slightly higher for the parental strain compared to the BG27 Δppk1 mutant, the addition of exogenous polyP did not change the outcome. Furthermore, the percentage of internalized bacteria was comparable among the wild-type BG27, the BG27 Δppk1 mutant strain and the condition encompassing the exogenous polyP addition (Fig 9C). Altogether these results pointed out a major involvement of polyP in the protection from intracellular killing by dHL60, suggesting that the polyP capsule could hinder the ability of phagocytes to kill internalized N. gonorrhoeae.

Discussion
PolyP is present in all living cells of both prokaryotes and eukaryotes where it displays multiple functions, including energy storage and metal ion chelator [24,[36][37][38]. In pathogenic bacteria,

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule polyP may also act as a virulence factor, contributing to motility, resistance to stress conditions, heat and acid tolerance or colonization [28,39,40]. In some bacterial species, including N. gonorrhoeae, polyP may be also localized even on the exterior of the cell. Although data remained still controversial, it has been suggested that its massive extracellular accumulation generated a capsule-like coating [13]. In this study, we demonstrate the existence of a polyP pseudo-capsule in gonococcus and propose its functional role in the evasion from human innate immune system. Utilizing the recombinant C-terminal domain of E. coli polyP-binding protein (EcPPXc), our experiments provided evidences of a pseudo-capsule and a bacteriaembedding matrix consisting of polyP in the reference strain FA1090. In gonococcus, the metabolism of polyP is mainly regulated by two polyphosphate kinases (PPK1 and PPK2) and one exopolyphosphatase (PPX). As already shown by Tinsley and Gotschlich in 1995, the wildtype FA1090 strain differs from ppk mutants in the total amount of polyP accumulation during all growth phases [27]. We here show that these mutations also influence the polyP content located at the exterior of the cell. Indeed, flow cytometry and confocal microscopy analyses indicate: i) the presence of two distinct bacterial sub-populations in the parental strain, one of which exposing polyP on the bacterial surface; ii) the disappearance of this polyP-positive subpopulation when one or both ppk genes are silenced; iii) and a uniform polyP-positive population associated with the ppx deletion mutant. These experimental evidences allowed to exclude a compensatory effect of PPK expression in the Δppx mutant or of PPX in the Δppk mutants, as observed in N. meningitidis by Zhang and colleagues [30]. However, other pleiotropic effects upon these mutations may not be excluded. The biological relevance of the presence of a polyP pseudo-capsule is highlighted by analyzing BG27 strain, a low-passaged clinical isolate harboring the PorB.1B, with polyP homogenously present on the surface of the whole population. In this work we show that a sustained external polyP content correlates with serum resistance, given that gonococcal mutants in which ppk genes have been inactivated are highly sensitive to complement-mediated lysis, whereas bacteria lacking PPX and presenting polyP on the external surface, exhibit a resistant phenotype. It is worth underlining that the aforementioned results with the clinical isolate BG27, indicating a resistant phenotype, clearly mirrored those obtained with FA1090 Δppx, suggesting that in the BG27 strain the amount of surface-associated polyP is naturally higher than that of FA1090 wild-type strain. To gain further insights into the basis of polyP-mediated serum resistance, we examined the deposition of C9 on bacteria following incubation with human complement. Complement deposition was tested in the presence of 10% NHS since it has been reported that complement factors account for a total amount of 11,5% of undiluted human serum in human cervical mucus [41]. In these experiments, we found an inverse correlation between polyP surface-exposure and complement deposition, strongly suggesting that external polyP may act as a physical-chemical barrier against the activation of the complement cascade on the bacterial membrane. Zhang and colleagues, whose findings on N. meningitidis are consistent with those observed in this study, proposed a major involvement of polyP in the regulation of the complement alternative cascade since the inhibition of the classical and lectin pathways still correlates with a significant survival advantage for the ppx mutant compared to the wild-type [30]. It is acknowledged that antibodies against N. gonorrhoeae, such as IgG and IgM elicited against PorB and LOS respectively, are present in serum of individuals with no history of gonococcal infections [42,43]. Since we did not remove immunoglobulins from sera used to assess survival and complement deposition on the different strains used in this study, we cannot completely exclude any additional role of polyP in the interference with the classical pathway. However, our studies on FA1090 and BG27 strains suggest a new mechanism used by N. gonorrhoeae to evade human serum bactericidal activity, an aspect that to date has not been fully elucidated, given that some strains harboring PorB.1B have a serum resistant phenotype in the absence of the binding ability towards both factor H and C4b-binding protein (C4bp) [4,44]. Results obtained using PorB.1B serum sensitive strain F62 seem to corroborate this newly proposed resistance mechanism for gonococcus. Although the unstable serum resistance that characterizes F62 may be ascribed to the inability of this strain to bind C4bp through its porin [44], the absence of a polyP pseudo-capsule shown in the present study introduces a new possible explanation for its in vitro sensitivity. The fact that exogenous polyP can be used to coat bacteria that do not naturally produce a pseudo-capsule, suggests that gonococcus might exploit free polyP released by human tissues to escape host complement-mediated killing mechanisms [30,45]. The observed survival rate of F62 strain in normal human serum adjunct of exogenous polyP corroborates this hypothesis, that is further complemented by the negligible C3 deposition on bacterial surface in the presence of externally added polyP. This hypothesis is also in accordance with in vivo gonococcal sialylation [46] as our results point up the contribution of polyP pseudo-capsule in the protection from bactericidal activity of normal human serum from different sources also in conditions promoting LOS sialylation. To further explore the polyP-dependent serum resistance the analysis was extended to PorB.1A bearing gonococci. In our experimental settings polyP pseudo-capsule was not detectable on the surface of the laboratory strain SK92-679 and of the clinical isolates WHO-N and WHO-F (S4 Fig). The mechanisms through which PorB.1A strains resist the complement-mediated killing are extensively described in literature and involve the ability of porin to bind factor H and C4bp, thus inhibiting the alternative and classical pathway, respectively [4,44,47,48]. To support this hypothesis, we showed that the addition of exogenous polyP does not give a survival advantage to the gonococci (S5 Fig) since resistance to high concentration of NHS is already conferred by porin.
The idea of a polyP pseudo-capsule as a protective shield led us to investigate its potential involvement in the protection from cationic antimicrobial peptides, another effective mechanism of human mucosal immune response against bacterial pathogens. In particular, LL-37, present in neutrophil granules and secreted by epithelial cells, exerts a potent bactericidal activity against gonococcus [49]. In the present study we show that LL-37 lethality increases with a reduced polyP pseudo-capsule production independently of LOS sialylation, whereas it is attenuated or even abrogated after the addition of exogenous polyP. Interestingly, LL-37 peptide can be found on mucosal sites at a concentration of about 0.44 μM [50] similarly to the concentration range where bactericidal activity against polyP negative strains was observed, highlighting the in vivo relevance of polyP-dependent defense mechanism against LL-37.
Another crucial aspect in gonococcal pathogenesis is the interaction with human polymorphonuclear leukocytes (PMNs) within which the bacterium is able to survive and replicate [3,6]. The mechanism by which gonococcus survives PMNs challenge still remains an overarching question. In this study, we provide evidences that polyP seems not to be directly involved in the phagocytic process but it is able to confer protection to intracellular killing by neutrophil-like HL60 cells, a validated cell culture model for N. gonorrhoeae-neutrophils interaction [51]. BG27 Δppk1 mutant, lacking a polyP pseudo-capsule has a significantly reduced survival in differentiated HL60 cells compared to its parental strain, results that are in agreement with the study of Sureka and colleagues on Mycobacterium tuberculosis, in which ppk1 attenuation compromised the ability to survive in macrophages [52]. Nevertheless, how polyP protects from killing in the aforementioned evasion mechanisms remained unclear. Recently, Rijal and co-workers proposed that polyP secreted by Mycobacterium smegmatis can act as a molecular signal to inhibit phagosome acidification and lysosome activity in macrophages to potentiate viability of ingested bacteria, suggesting a possible explanation for the decreased pathogenicity of the bacteria lacking polyP pseudo-capsule [53].
In conclusion, results of this study provide clear evidence that the presence of a polyP pseudo-capsule in N. gonorrhoeae confers protection against different human immune response mechanisms. However, the different phenotypes observed using different gonococcal strains cannot be explained by the presence of non-functional catalytic sites (S6 Fig), or relevant differences in the expression levels of the enzymes involved in polyP turnover (S7 Fig). These findings were not surprising, since polyP can first function in gonococcus as an energy storage pool. Other gene products may be possibly involved in polyP translocation across the outer membrane and may be differentially regulated in different strains. A multi-omic approach would be beneficial to better clarify the mechanism regulating the production of gonococcal polyP pseudo-capsule. Although further studies are needed, our results introduce a further strategy applied by gonococcus to evade the innate immune response of the host, representing a breakthrough in the field of gonococcal biology and paving the way to a new perspective for prevention strategies and treatments, as well as for new research on gonococcal patho-biological mechanisms [54].

Ethics statement
Human complement source used in the study was obtained according to Good Clinical Practice in accordance with the declaration of Helsinki. Patients have given their written consent for the use of samples of study MENB REC 2ND GEN-074 (V72_92). The study was approved by the Western Institutional Review Board (WIRB).

Bacterial strains and growth conditions
N. gonorrhoeae and Escherichia coli strains used in this work are listed in Table 1. N. gonorrhoeae was grown in gonococcal base (GC) agar plates containing 1% v/v Isovitalex at 37˚C in an atmosphere of 5% CO 2 . For liquid cultures, bacteria grown for at least 18 h on the plates, were diluted to an OD 600 of 0.1 in 10 mL of liquid GC-1% v/v Isovitalex or in the chemical defined medium described by Wong and colleagues (WSJM medium) [55] and incubated at 37˚C at 160 rpm. To stimulate sialylation of F62, when necessary, the cytidine 5'-monophospho-N-acetylneuraminic acid (CMP-NANA) was added to the liquid medium in a concentration of 0.5 mg/L, 50 mg/L or 500 mg/L. When required, kanamycin (60 mg/L for FA1090 and 80 mg/L for BG27) and chloramphenicol (2 mg/L) were added to culture media at the indicated final concentrations.
E. coli was grown either in Luria Bertani (LB) liquid medium with shaking at 180 rpm or on LB agar plates overnight at 37˚C for plasmid purification or in High Throughput Medium Complex (HTMC) (15 g/L glycerol, 30 g/L yeast extract, 0.5 g/L MgSO 4 , 5 g/L KH 2 PO 4 , 20 g/L K 2 HPO 4 ) at 180 rpm at 25˚C for protein purification. When required, kanamycin, chloramphenicol or ampicillin were added to achieve a final concentration of 30 mg/L, 20mg/L and 100 mg/L respectively.

HL60 cell culture and differentiation
The HL60 cell line was obtained from ATCC (American Type Culture Collection). Undifferentiated HL60 cells were cultured in RPMI 1640 with GlutaMAX supplement (Gibco), containing 10% v/v heat-inactivated fetal bovine serum (GE Healthcare) and were kept at concentrations between 10 5 and 10 6 cells/mL. Differentiation of HL60 to neutrophil-like cells was achieved by diluting cells to 4 x 10 5 cells/mL, followed by treatment with 0.78% v/v dimethylformamide (Sigma Aldrich) in culture medium for 4 to 6 days, accordingly to Romero-Steiner et al. [56].

Construction of plasmids for generation of N. gonorrhoeae mutants
Genomic DNA and plasmid DNA purification was performed using GenElute Bacterial Genomic DNA kit (Sigma Aldrich) and QIAquick PCR Purification kit (Qiagen) according to the manufacturer's instruction. Plasmids used in this study are listed in Table 2. To generate the constructs used to knockout the ppk1 (NGO 0003) and ppk2 (NGO 2113) genes in N.

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule gonorrhoeae, Polymerase Incomplete Primer Extension (PIPE) cloning was applied [57]. Briefly, gene loci including the coding sequence and 500 bp upstream and downstream sequences, have been amplified from FA1090 genomic DNA and cloned in modified pET-15 plasmid, using ppkKOupF/ppkKOdoR and ppk2KOupF/ppk2KOdoR, respectively (Table 3). In a second step the coding sequence was replaced by kanamycin and chloramphenicol resistance cassette [57] respectively previously amplified from pET-24 (Novagen) and pSLComCmr [58] with KanKOF/KanKOR and CloKOF/CloKOR primers. To generate construct used to knockout ppx (NGO 1041) in N. gonorrhoeae, the restriction enzymes-based cloning was performed: 500 bp of downstream and 500 bp upstream fragments were individually amplified using primers ppxUPfXBA/ppxUPrsma and ppxDOfsma/ppxDOrXHO and digested with XbaI-SmaI and SmaI-XhoI respectively. In parallel, the chloramphenicol resistance cassette was PCR amplified from pSLComCmr [58] with CloSMAF/CloSMAR and digested with SmaI restriction enzyme. The resulting fragments were ligated into the pBlueScript (Agilent) plasmid previously digested with XhoI and XbaI restriction enzymes. All the PCR amplifications were performed using the KAPA Hi-FI polymerase (KAPA Biosystem). The restriction enzymes digestion was performed following New England Biolabs manual and the ligation with T4 DNA ligase from Thermo Fisher Scientific. DNA samples were analyzed by agarose gel electrophoresis and visualized by staining with SYBR Safe (Thermo Fisher Scientific). E. coli Mach-1 and DH5-α competent cells were respectively used to perform PIPE and restriction enzyme based cloning steps and transformed by heat shock, according to the Thermo Fisher Scientific protocol. To transform N. gonorrhoeae, strains were grown on agar plates for at least 18 h. A swab of bacteria was resuspended in PBS, mixed with purified DNA (*1 μg) and then spotted on GC-1% v/v Isovitalex agar plates. The bacteria were then incubated for 5 h in the presence of 5% CO 2 at 37˚C and then finally plated to agar plates containing the appropriate antibiotic. To efficiently screen the transformants, single colonies were expanded onto agar plates in the presence of the appropriate antibiotic and incubated for 18 h at 37˚C in an atmosphere of 5% CO 2 . A loopful of bacteria was resuspended in PBS and lysed by incubating the bacterial suspension at 100˚C for 10 min in a thermo mixer. After a centrifugation step at 13000 rpm for 5 min, the resulting supernatant containing DNA was used for PCR screening with AccuPrime DNA Polymerase (Thermo Fisher Scientific) according to the manufacturer's protocol. External oligonucleotides (Table 3) were employed to check if recombinations occurred in the expected genome locus.

Transmission electron microscopy
N. gonorrhoeae FA1090 colonies grown overnight at 37˚C in the presence of 5% CO 2 on GC-1% v/v Isovitalex plates, were suspended to 0.1 OD 600 in 50 mL GC-1% v/v Isovitalex liquid broth prewarmed to 37˚C and incubated until the log phase (0.5 OD 600 ). The cell suspension was harvested at 4000 rpm for 5 min. Samples were immersed in 2.5% v/v glutaraldehyde, 2% v/v paraformaldehyde in 0.1 M cacodylate sucrose buffer pH 7.2, containing 0.15% v/v Alcian blue and 1.55% w/v L-lysine acetate, for 20 min at 4˚C. This was followed by the primary fixative alone (without L-lysine-acetate), for 3 h at 4˚C. For the negative control the L-lysine acetate was not included in the solutions. The fixed cells were then washed in the same buffer (three rinses of min each), and post-fixed with 1% osmium tetroxide in 0.1 M cacodylate buffer containing 0.15% v/v Alcian blue for 1 h at 4˚C. Specimens were washed three times, 15 min each, at 4˚C in the same buffer, dehydrated in a graded ethanol series, then infiltrated with mixtures of LRWhite resin/ethanol in different percentages. At the end of the procedure, samples were embedded in LRWhite resin, 2 days at 50˚C. Blocks were cut with Reichert Ultracut ultramicrotome using a diamond knife. For TEM, Ultrathin sections (60-80 nm) were

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule collected on copper grids, stained with uranyl acetate and lead citrate, and observed with a JEOL 1200 EX II electron microscope. Micrographs were captured by the Olympus SIS VELETA CCD camera equipped with iTEM software.

Production and purification of a recombinant His-tagged C-terminal domain of the E. coli exopolyphosphatase
The enzymatically inactive C-terminal domain of the E. coli strain K12 exopolyphosphatase (EcPPXc), that is a specifically polyP binding domain [31], was amplified and cloned in the pET 15 plasmid. E. coli K12 genomic DNA was used as DNA template for the EcPPXc gene amplification. The PCR product amplified using primers Ecppx_F and Ecppx_F (Table 3) was cloned in a pET 15 plasmid applying a PIPE cloning strategy, thus obtaining the pET 15-EcPPXc. The plasmid was then transformed into E. coli BL21 (DE3). Single colonies of recombinant E. coli were inoculated in 5 mL of LB added with 100 mg/L ampicillin and grown overnight at 37˚C, 180 rpm. The cultures were then diluted 1:100 in 100 mL of HTMC medium with 100 mg/L ampicillin. Protein expression was induced with isopropyl D-thiogalactoside (IPTG) at a final concentration of 1 mM after approximatively 6 h growth at 25˚C, 180 rpm. Then the culture was incubated overnight until stationary phase. For protein purification, the bacterial pellet was resuspended in the equilibration buffer (Tris-HCl 20 mM pH 8, NaCl 300 mM) and mechanically lysed by sonication (amplitude 65%, 30 sec ON/OFF, 30 min in total). Bacterial debris were removed by centrifugation and the recombinant protein was purified using the N-terminal His-tag on a cobalt-based chelating column (HiTrap TALON crude, GE Healthcare Life Science) by Immobilized Metal Affinity Chromatography (IMAC). The column was washed with equilibration buffer and proteins bound to the resin were then eluted with imidazole gradient (from 0 mM to 300 mM imidazole) in equilibration buffer. After this step, sample purity was assessed on an analytical size-

Staining of polyP on bacterial surface by Fluorescence Activator Cell Sorting (FACS)
N. gonorrhoeae colonies grown overnight at 37˚C in the presence of 5% CO 2 on GC-1% v/v Isovitalex plates, were suspended to 0.1 OD 600 in 10 mL GC-1% v/v Isovitalex broth liquid prewarmed to 37˚C and incubated until exponential phase at 37˚C at 160 rpm. Cell suspensions were sub-cultured into Wong medium and followed until early log phase (0.3-0.4 OD 600 ). A volume of bacterial culture was collected to have 1 OD 600 , spun down and fixed for 1 h at RT in 1 mL of 0.5% v/v formaldehyde in PBS. The bacterial pellet was then resuspended for 1 h at RT in blocking buffer (1% w/v BSA in PBS) to reach 0.25 OD 600 . PolyP staining was carried out by incubating for 30 min at RT 50 μL (OD 600 0.0125) of the cell suspension with 50 μL of blocking buffer containing the recombinant protein EcPPXc, to reach a final concentration of 100 μg/mL. As a negative control the cell suspension is incubated with the same volume of blocking buffer. The samples were washed once and incubated for 1 h with 100 μL primary antibody (1:250 monoclonal anti 6x-His Tag antibody, Thermo Fisher Scientific). After two washes, 100 μL of FITC-conjugated goat anti-mouse IgG secondary antibody was added at a 1:500 dilution and incubated for 30 min (Sigma Aldrich). After final washings, cells were resuspended in PBS. Bacterial fluorescence was recorded with BD FACS CANTO II (BD Bioscience) and data were analyzed using Flow-Jo v.10 (FloJo, LLC).

Serum resistance assay
Gonococci were grown to the log phase (OD 600 , between 0.4 and 0.5) in GC-1% v/v Isovitalex medium, subcultured into Wong medium and then followed until early exponential phase (0.3-0.4 OD 600 ). Fifty microliters of bacterial suspension, diluted 1:50 in a sterile buffer containing 1% w/v BSA 0,1% w/v glucose in PBS was incubated for 1 h at 37˚C in the same volume of 40% v/v and 20% v/v freshly thawed NHS, in order to assess bacteria survival in the presence of 20% v/v and 10% v/v of complement source, respectively. For competition experiment 50 μM of polyP (Graham's salt, Merck) or 50 μM of DNA (deoxyribonucleic acid, low molecular weight from salmon sperm, Sigma Aldrich) were added to the reaction mixture. Experimental controls included bacteria incubated with heat-inactivated complement at 56˚C for 30 min. Aliquots (10 μL) of diluted bacterial suspension were collected at the beginning of the experiment for enumeration of cfu at time zero (T 0 ) and after the incubation time to be disseminated into GC-1% v/v Isovitalex agar plates by tilt method.

Complement binding studies
For complement deposition studies, gonococci, grown as described above, were fixed in 0.5% v/v formaldehyde for 1 h, then resuspended in blocking buffer and finally 10 5 cells were incubated with 10% v/v NHS (or heat-inactivated NHS) for 30 min at 37˚C. For competition experiments, 50 μM of polyP (Graham's salt, Merck) was added to the reaction mixture. After two washes with 100 μL PBS, the samples were incubated for 1 h with 100 μL FITC-conjugated mouse anti human/mouse C3/C3b/iC3b antibody (1:100 dilution; Caderlane) or mouse antihuman C9 monoclonal antibody (at a 1:250 dilution; Thermo Fisher Scientific). Following washing steps, 100 μL of FITC-conjugated goat anti-mouse IgG secondary antibody was added at a 1:500 dilution and incubated for 30 min (Sigma Aldrich). Finally, cells were resuspended in PBS and bacterial fluorescence was recorded with BD FACS CANTO II (BD Bioscience) and then analyzed using Flow-Jo v.10 (FloJo, LLC).

Phagocytosis assay
Phagocytic uptake was assessed adapting a protocol applied by Chen and Seifert [51]. In a 24-well plate, 10 6 differentiated HL60 (dHL60) were seeded in a volume of 0.4 mL of infection medium consisting of RPMI 1640 with GlutaMAX supplement and were then allowed to adhere at 37˚C and 5% CO 2 for 30 min. The adherent dHL60 cells were then infected with 5 × 10 7 cfu of N. gonorrhoeae BG27 strain and BG27 Δppk1 mutant, grown in GC-1% v/v Isovitalex medium, in 0.1 mL volume of infection medium. In the condition with polyP, exogenous polyP was added to the infection medium to a final concentration of 50 μM polyP (Graham's salt, Merck). The infections were carried out at 37˚C and 5% CO 2 . At 1 h post infection, unbound bacteria were removed through a washing step with PBS. Cells were then lysed in 1% saponin in PBS to determine both adherent and internalized bacteria or were treated with 100

PLOS PATHOGENS
Gonococcal polyphosphate pseudo-capsule mg/L of gentamycin for 1 h to determine internalized bacteria only. Serial dilutions were then performed and plated on GC-1% v/v Isovitalex agar plates. The experiment was performed in triplicate while the complementation with free polyP in duplicate. Viable internalized bacteria were established as the ratio of phagocytized bacteria to total bacteria (adherent plus internalized bacteria). Statistical significance was determined with a one-way ANOVA, using Graph-Pad Prism 7 software (*p < 0.05, **p < 0.005). water objective was used to analyze the samples. In particular, a total of 25 z-stacks per field of acquisition (0.5 μm distance between planes) and 81 fields per well were acquired in confocal mode. Images were then analyzed in 3D mode by using the Harmony software. In particular, cell nuclei and cytoplasm were identified with DAPI and CellMask DeepRed fluorescence (S9 Fig). Afterwards the fluorescence volume of green-stained (total bacteria volume) and redstained (external bacteria volume) were calculated and used as reference of bacterial biomass.

Fluorescent-based phagocytosis assay
Using these values, the relative amount of internalized and external bacteria was finally calculated. The statistical significance was determined by two-way ANOVA test (*p < 0.05).

Antimicrobial assay
Bacteria grown in GC-1% v/v Isovitalex medium until early exponential phase, were suspended in the same liquid medium to a concentration of 10 5 cfu/mL. To assess the antimicrobial activity of the peptide LL-37 (Sigma Aldrich) 180 μL of bacterial suspension were added to each well of a sterile 96-well plate containing 20 μL of peptide solution at several concentrations that ranged from 50 μM to 1.5 μM, in order to have a final concentration from 5 μM to 0.15 μM. In experiments with the Human Neutrophil Peptide-1 (HNP-1, Sigma Aldrich) and the Human β-Defensin-1 (HBD-1, Sigma Aldrich), 25 μL of bacterial suspension were added to the same volume of peptide, suspended in order to have final peptide concentrations ranging from 5 μM to 0.078 μM and from 2.4 μM to 0.075 μM, respectively. The samples were then incubated at 37˚C for 1 h. In competition experiments, 20 μM polyP (Graham's salt, Merck) or 20 μM DNA (deoxyribonucleic acid, low molecular weight from salmon sperm, Sigma Aldrich) were added to the incubation mixture. The activity of the peptides was represented as the minimum bactericidal concentration (MBC), defined as the lowest peptide concentration where bacterial growth was not detected. MBC values were determined by plating serial dilutions of the reaction mixture on GC-1% v/v Isovitalex agar plates and enumerating the cfu on day after. The percentage of surviving bacteria was expressed as cfu after peptide incubation divided by cfu after control incubation and multiplied by 100, according to Ong and colleagues [60].